Examining Personal Lifestyle Factors and Lipid Profile on Cardiovascular Disease Risk Prediction

Cardiovascular disease (CVD) is a major global health challenge, with various factors such as lifestyle choices, lipid profiles, and genetic predispositions contributing to its development. Early and accurate prediction of CVD risk is critical for effective prevention and management strategies. This study aims to develop robust predictive models for CVD risk by integrating demographic, health, genetic, and imaging data, utilizing machine learning (ML) and deep learning (DL). The primary objective is to leverage X-ray image data analyzed through convolutional neural networks (CNNs) to improve the accuracy of risk prediction models. A secondary objective is to explore the potential of combining structured patient data, including demographic information, lipid profiles, and genetic markers, with advanced imaging features to provide a comprehensive and precise CVD risk assessment. By integrating ML and DL algorithms, the study also aims to identify key features and patterns that could contribute to early detection and personalized healthcare interventions for individuals at risk of CVD. The dataset used in this study consists of 264 patients, incorporating demographic details, health indicators, lipid profiles, genetic markers, lifestyle factors, and X-ray images to detect heart-related abnormalities. Data preprocessing techniques such as normalization, resizing, and augmentation were applied to ensure consistency across the dataset. Feature selection was performed using techniques like LASSO and correlation analysis to identify the most predictive variables. Various ML models, including Support Vector Machines (SVM), Random Forest (RF), and Decision Trees (DT), were trained on the curated data. CNNs were employed for image data analysis, and Grad-CAM was used to visualize and interpret the model's decision-making process. The results showed promising outcomes, with ML models achieving up to 95% accuracy, and CNNs using X-ray images achieving 100% accuracy. This research highlights the potential of integrating ML and DL algorithms for early detection and personalized healthcare strategies in CVD.